Head and Neck

Tissue Components and Segmentation of Early Craniofacial Region

The early craniofacial region consists of a massive neural tube beneath which lie the notochord and a ventrally situated pharynx (see Figure 14.2 ). The pharynx is surrounded by a series of pharyngeal arches. Many of the tissue components of the head and neck are organized segmentally. Figure 14.3 illustrates the segmentation of the tissue components of the head. As discussed in earlier chapters, distinct patterns of expression of certain homeobox-containing genes are associated with morphological segmentation in some cranial tissues, particularly the central nervous system (see Figure 11.14 ). The chain of events between segmental patterns of gene expression and the appearance of morphological segmentation in parts of the cranial region is the subject of much current research.

Fundamental Organization of the Pharyngeal Region

Because many components of the face are derived from the pharyngeal region, an understanding of the basic organization of this region is important. In a 1-month-old embryo, the pharyngeal part of the foregut contains four lateral pairs of endodermally lined outpocketings called pharyngeal pouches and an unpaired ventral midline diverticulum, the thyroid primordium ( Figure 14.2B ). If the contours of the ectodermal covering over the pharyngeal region are followed, bilateral pairs of inpocketings called pharyngeal grooves , which almost contact the lateralmost extent of the pharyngeal pouches, are seen.

Alternating with the pharyngeal grooves and pouches are paired masses of mesenchyme called pharyngeal (branchial) arches . Central to each pharyngeal arch is a prominent artery called an aortic arch , which extends from the ventral to the dorsal aorta (see Chapter 17 and Figure 14.2B ). The mesenchyme of the pharyngeal arches is of dual origin. The mesenchyme of the incipient musculature originates from mesoderm, specifically the somitomeres. Much of the remaining pharyngeal arch mesenchyme, especially that of the ventral part, is derived from the neural crest, whereas mesoderm makes various contributions to the dorsal pharyngeal arch mesenchyme.

Cellular Migrations and Tissue Displacements in the Craniofacial Region

Early craniofacial development is characterized by several massive migrations and displacements of cells and tissues. The neural crest is the first tissue to exhibit such migratory behavior, with cells migrating from the nervous system even before closure of the cranial neural tube (see Chapter 12 ). Three streams of neural crest cells migrate into the craniopharyngeal region (see Figure 12.7 ). The first, arising from the area of the first two rhombomeres, streams into the future face and jaws (first pharyngeal arch). The second, arising from the area of rhombomere 4, leads into the second pharyngeal arch. Cells of the third stream, arising from several rhombomeric regions, populate the remaining pharyngeal arches.

Initially, segmental groups of neural crest cells are segregated, especially in the pharyngeal region (see Figure 14.3 ). These populations of cells become confluent, however, during their migrations through the pharyngeal arches. Much of the detailed anatomy of the facial skeleton and musculature is based on the timing, location, and interactions of individual streams of neural crest and mesodermal cells. Recognition of this level of detail (which is beyond the scope of this text) is important in understanding the basis underlying many of the numerous varieties of facial clefts that are seen in pediatric surgical clinics.

The early cranial mesoderm consists mainly of the paraxial and prechordal mesoderm (see Figure 14.3 ). Although the paraxial mesoderm rostral to the occipital somites has often been considered to be subdivided into somitomeres, many embryologists now classify it as being unsegmented mesoderm (see Figure 14.3 ). Mesenchymal cells originating in the paraxial mesoderm form the connective tissue and skeletal elements of the caudal part of the cranium and the dorsal part of the neck. Within the pharyngeal arches, cells from the paraxial mesoderm initially form a mesodermal core, which is surrounded by cranial neural crest cells (see Figure 14.4A ). At this stage of development, the two types of cells do not intermix. Myogenic cells from paraxial mesoderm undergo extensive migrations to form the bulk of the muscles of the cranial region. Similar to their counterparts in the trunk and limbs, these myogenic cells become integrated with local connective tissue to form muscles. Another similarity with the trunk musculature is that morphogenetic control resides within the connective tissue elements of the muscles, rather than in the myogenic cells themselves. In the face and ventral pharynx, this connective tissue is of neural crest origin.

The prechordal mesoderm, which emits important forebrain inductive signals in the early embryo, is a transient mass of cells located in the midline, rostral to the tip of the notochord. Although the fate of these cells is controversial, some investigators believe that myoblasts contributing to the extraocular muscles take origin from these cells. On their way to the eye, cells of the prechordal mesoderm may pass through the rostralmost paraxial mesoderm.

The lateral plate mesoderm is not well defined in the cranial region. Transplantation experiments have shown that it gives rise to endothelial cells, smooth and skeletal muscle (see p. 190), and, at least in birds, to some portions of the laryngeal cartilages.

Another set of tissue displacements important in the cranial region is the joining of cells derived from the ectodermal placodes with cells of the neural crest to form parts of sense organs and ganglia of certain cranial nerves (see Figure 13.1 ).

Formation of the Face and Jaws

Development of the face and jaws is a complex three-dimensional process involving the patterning, outgrowth, fusion, and molding of various tissue masses. The forebrain acts as a mechanical substrate and a signaling center for early facial development, and the stomodeum serves as a morphological point of reference. The lower face (maxillary region and lower jaw) is phylogenetically derived from a greatly expanded first pharyngeal arch. Much of the mesenchyme of the face is neural crest, originating from the forebrain to the first two rhombomeres. Each of the tissue components of the early face is the product of a unique set of morphogenetic determinants and growth signals, and increasing evidence indicates that specific sets of molecular signals control their development along the proximodistal and the rostrocaudal axes.

At a higher level, the building blocks of the face relate to one another in highly specific ways, and clues to their origins and relationships can be derived by examination of their blood supply. Disturbances at this level regularly result in the production of craniofacial anomalies, and an understanding of the fundamental elements of facial morphogenesis is crucial to rational surgical approaches to these malformations.

Structures of the face and jaws originate from several primordia that surround the stomodeal depression of a 4- to 5-week-old human embryo ( Figure 14.5 ). These primordia consist of the following: an unpaired frontonasal prominence ; paired nasomedial and nasolateral processes , which are components of the horseshoe-shaped olfactory (nasal) primordia; and paired maxillary processes and mandibular prominences , both components of the first pharyngeal arches. The olfactory primordia form around the olfactory placodes, which are present as focal ectodermal thickenings on the ventrolateral sides of the frontonasal prominence.

The upper jaw contains a mixed population of neural crest cells derived from the forebrain and midbrain, whereas the lower jaw contains mesenchymal cells derived from midbrain and hindbrain (rhombomeres one and two) neural crest. The specific morphology of facial skeletal elements is determined by signals passed from the pharyngeal endoderm to the facial ectoderm and then to the neural crest precursors of the facial bones. Narrow zones of pharyngeal endoderm control the morphogenesis of specific portions of the skeleton of the lower face. FGF-8 signaling from the facial ectoderm plays a key role in patterning of the facial skeleton.

Another factor that strongly influences facial form is the responsiveness of the various facial processes to Wnts . In many developing structures, Wnt signaling stimulates cellular proliferation that increases the mass of that structure. In facial development, species with an elongated midface (e.g., the beak of birds) have a midline zone of Wnt responsiveness within the frontonasal process. Other species (e.g., humans) that have flat but broad faces have Wnt-responsive regions in the maxillary and mandibular processes, thus supporting lateral growth in the face.

The frontonasal prominence is a dominant structure in the earliest phases of facial development, and its formation is the result of an exquisitely sensitive signaling system that begins with the synthesis of retinoic acid in a localized region of ectoderm opposite the forebrain and continues with the action of shh produced by the ventral forebrain. The action of shh, through the mediation of the most rostral wave of neural crest cells, underlies the establishment of the frontonasal ectodermal zone, located at the tips of the nasomedial processes (see p. 300). The signaling molecules (FGF-8 and shh) emanating from this zone stimulate cell proliferation in the neural crest mesenchyme of the frontonasal prominence. The same signaling molecules also stimulate proliferation of the cells of the maxillary process as it begins to form the basis for the upper jaw. In the absence of signaling from the frontonasal ectodermal zone, cell death in the region increases, and cell proliferation decreases, resulting in various midfacial defects (see Clinical Correlation 14.1 ). Retinoic acid is unusual in that both deficiencies and excess amounts can cause very similar defects. From 4 to 5 weeks, the frontonasal prominence is a dominant structure of the early face (see Figure 14.5 ), but with subsequent growth of the maxillary process and the nasomedial and nasolateral processes, it recedes from the oral region. The nasolateral process develops as a result of FGF signaling emanating from the nasal pit and stimulating proliferation of mesenchymal cells in the nasolateral process.

Organization of the pharyngeal arches can be viewed from several different perspective s . Along the craniocaudal axis, the first arch and its derivatives function independently of Hox gene control, whereas arches 2–4 are specified by the Hox code, with the paralog Hox-2 dominating the second arch, Hox-3 , the third arch, and Hox-4 , the fourth arch. Organization along the dorsoventral axis is represented most prominently by graded expression of DLX genes, which themselves operate downstream from localized molecular signals and are upstream of other morphogenetic effector molecules. Details of molecular control along the dorsoventral axis varies considerably, especially among the more cranial components of the pharyngeal arch system. Even structures along the mediolateral axis, especially in the first arch, develop under the influence of morphogenetic gradients.

The maxillary and mandibular processes have traditionally been considered derivatives of the first pharyngeal arch. More recent research suggests that although some of the cells that form the maxillary process arise from the first arch, many mesenchymal cells of the maxillary process are not first-arch derivatives, but instead come from other areas of cranial neural crest. How these cells are integrated into a unified structure and what controls their specific morphogenesis remain to be determined.

The subdivision of the first arch into maxillary and mandibular regions is controlled to a large extent by endothelin-1 . Expressed at the distal (ventral) tip of the arch, endothelin-1 is required for formation of the mandibular process. Endothelin-1 promotes the expression of distal genes, such as Dlx-3,4, -5/6, and their downstream targets ( Hand-2 and Goosecoid ), which pattern the mandible. At an intermediate dorsoventral level within the first arch, endothelin-1 stimulates the early expression of Barx-1 , which is a prime determinant of the formation of the mandibular joint. When endothelin-1 is mutated or inactivated, the mandible becomes transformed into a structure resembling the maxilla. If endothelin-1 is overexpressed in the proximal part of the first arch, the future maxilla becomes transformed into a mandible. This effect is transmitted through the activation of Dlx-5/6. In the proximal (dorsal) part of the first arch, the influence of endothelin-1 is reduced, and active patterning genes (e.g., Dlx-1/2 ) lay the foundation for the formation of both the maxilla and the middle ear bones (malleus, incus, and tympanic ring).

Despite the relatively featureless appearance of the early mandibular process (first pharyngeal arch), the mediolateral (oral-aboral), and proximodistal axes are tightly specified ( Figure 14.6 ). This recognition has considerable clinical significance because increasing numbers of genetic mutants are recognized to affect only certain regions of the arch, such as the absence of distal (adult midline) versus proximal structures. The medial (oral) region of the mandibular process, which seems to be the driver of mandibular growth, responds to local epithelial signals (FGF-8) by stimulating proliferation of the underlying mesenchyme through the mediation of Msx-1, similar to the subectodermal region of the limb bud. Growth of the jaws is influenced by various growth factors, especially the bone morphogenetic proteins (BMPs), which, at different stages, are produced in either the ectoderm or the mesenchyme and can have strikingly different effects. Experiments on avian embryos have shown that increasing the expression of BMP-4 in first-arch mesenchyme results in the formation of a much more massive beak than that of normal embryos. Additional experiments on avian embryos have shown that much of the morphogenetic information required to form facial structures is carried in the neural crest mesenchyme.

Proximodistal (dorsoventral) organization of the first arch is reflected by nested expression patterns of the transcription factor Dlx (the mammalian equivalent of distalless in Drosophila ) along the arch (see Figure 14.6A ). Opposing gradients of FGF-8 (distal) and BMP-4 (proximal) set the overall proximodistal (dorsoventral) pattern and restrict the expression of Dlx-1,-2 and Barx-1 to the more proximal regions of the first arch. Dlx-1,-2 expression specifies the maxillary process, whereas Dlx-1,-2 and Dlx-5,-6 specify the mandibular process. Within the mandibular process, Dlx-5,-6 is expressed slightly more distally than Dlx-1,-2 . It also controls the more distal expression of Dlx-3,-4 . Hand-2 is expressed distally in the early mandibular arch. Hand-2 represses osteogenesis in the distal lower jaw, thus preparing the way toward formation of the tongue.

Although mutants of individual Dlx genes produce minor abnormalities, mice in which Dlx-5 and Dlx-6 have been knocked out develop with a homeotic transformation of the distal lower jaws into upper jaws. Lhx-1 (rostral) and Gsc (caudal) expression sets the molecular boundaries for the rostrocaudal axis of the mandibular process. Within the oral-aboral axis, expression of Pax-3 sets the boundary between the future tooth-forming area (alveolar ridge) and the tongue.

Through differential growth between 4 and 8 weeks (see Figure 14.5 ), the nasomedial and maxillary processes become more prominent and ultimately fuse to form the upper lip and jaw ( Figure 14.7 ). While this is occurring, the frontonasal prominence, which was a prominent tissue bordering the stomodeal area in the 4- and 5-week-old embryo, is displaced as the two nasomedial processes merge. The sites of the frontonasal ectodermal zones on the merged nasomedial processes mark the distalmost tip of the upper jaw. The merged nasomedial processes form the intermaxillary segment , which is a precursor for (1) the philtrum of the lip, (2) the premaxillary component of the upper jaw, and (3) the primary palate .

The maxillary processes also respond to the expansion of the early forebrain by being pushed laterally. This is an example of a physical process affecting shape. In general, a larger brain leads to a wider face.

Between the maxillary process and the nasal primordium (nasolateral process) is a nasolacrimal groove (nasooptic furrow) that extends to the developing eye (see Figure 14.5 ). The ectoderm of the floor of the nasolacrimal groove thickens to form a solid epithelial cord, which detaches from the groove. The epithelial cord undergoes canalization and forms the nasolacrimal duct and, near the eye, the lacrimal sac . The nasolacrimal duct extends from the medial corner of the eye to the nasal cavity (inferior meatus) and, in postnatal life, acts as a drain for lacrimal fluid. This connection explains why people can have a runny nose when crying. Meanwhile, the expanding nasomedial process fuses with the maxillary process, and over the region of the nasolacrimal groove, the nasolateral process merges with the superficial region of the maxillary process. The region of fusion of the nasomedial and maxillary processes is marked by an epithelial seam, called the nasal fin . Mesenchyme soon penetrates the nasal fin, and the result is a continuous union between the nasomedial and maxillary processes. The remaining part of the nasal fin goes on to form the oronasal membrane (see p. 310).

The lower jaw is formed in a simpler manner. The mandibular primordia are initially populated by cranial neural crest cells from the r1 and r2 levels. The bilateral mandibular prominences enlarge, and their medial components merge in the midline, to form the point of the lower jaw. The midline dimple that is seen in the lower jaw of some individuals reflects variation in the degree of merging of the mandibular prominences. A prominent cartilaginous rod called Meckel’s cartilage differentiates within the lower jaw (see Figure 14.40D ). Derived from neural crest cells of the first pharyngeal arch, Meckel’s cartilage forms the basis around which membrane bone (which forms the definitive skeleton of the lower jaw) is laid down. The developing Meckel’s cartilage can be subdivided into three main regions. The most proximal region gives rise to the primordia of the malleus of the middle ear. (The incus arises from the palatoquadrate cartilage, another element of the first arch.) Between the spine of the sphenoid and the malleus is a small Meckel’s cartilage remnant, the anterior ligament of the malleus. The middle portion degenerates and ultimately gives rise to the sphenomandibular ligament (see p. 328). The distal parts of both Meckel’s cartilages grow in length until their tips meet and fuse, forming a fibrous mandibular symphysis. Under the influence of Indian hedgehog, the fused tips undergo endochondral ossification while they form the definitive symphysis mentis during the second postnatal year. Formation of the definitive mandible does not depend upon Meckel’s cartilage, because the bony mandible can form in the absence of Meckel’s cartilage.

Experimental evidence indicates that the rodlike shape of Meckel’s cartilage is associated with the inhibition of further chondrogenesis by the surrounding ectoderm. If the ectoderm is removed around Meckel’s cartilage, large masses of cartilage form instead of a rod. These properties are similar to the inhibitory interactions between ectoderm and chondrogenesis in the limb bud. A later-acting influence in growth of the lower jaw is the planar cell polarity pathway, which influences outgrowth of Meckel’s cartilage. If this pathway is disrupted, the mandible will not develop to its normal length.

Shortly after the basic facial structures take shape, they are invaded by mesodermal cells associated with the first and second pharyngeal arches. These cells form the muscles of mastication (first-arch derivatives, which are innervated by cranial nerve V) and the muscles of facial expression (second-arch derivatives, which are innervated by cranial nerve VII). At the level of individual muscles, highly coordinated spatiotemporal relationships between mesodermal and neural crest cells are very important in the determination of muscle attachments and the overall shape of the muscles.

Although the basic structure of the face is established between 4 and 8 weeks, changes in the proportionality of the various regions continue until well after birth. In particular, the midface remains underdeveloped during embryogenesis and early postnatal life.

Temporomandibular Joint and Its Relationship With the Jaw Joint of Lower Vertebrates

Of considerable clinical importance and evolutionary interest is the temporomandibular joint , which represents the hinge between the mandibular condyle and the squamous part of the temporal bone. The temporomandibular joint, which phylogenetically appeared with the evolution of mammals, is a complex synovial joint surrounded by a capsule and containing an articular disk between the two bones. Within the joint, the articular disk is interposed between an upper and a lower synovial cavity. Based on the early expression of Barx-1, this joint is formed late during development, first appearing as mesenchymal condensations associated with the temporal bone and mandibular condyle during the seventh week of development. The articular disk and capsule begin to take shape a week later, and the actual joint cavity forms between weeks 9 and 11.

The cartilage of the temporomandibular joint differs from that of almost all other joints in that it is secondary cartilage . Secondary cartilage arises after, not before, actual bone formation by the conversion of periosteum in force-bearing areas to perichondrium. Its formation and presence is highly dependent upon function. Mechanical pressure fosters cartilage formation, whereas in the absence of function, the amount of cartilage is reduced, and additional bone is formed in the area. Corresponding to this dual potential, the cells of secondary cartilage express both SOX9 and RUNX2 (see Figure 9.20 ).

In lower vertebrates, the jaw opens and closes on a hinge between cartilaginous portions of the mandibular process—the articular bone in the lower jaw and the quadrate bone in the upper jaw. During phylogenesis, the distal membranous bone (the dentary bone ) associated with Meckel’s cartilage increased in prominence as the jaw musculature became more massive. The dentary bone of contemporary mammals and humans constitutes most of the lower jaw, and Meckel’s cartilage is seen only as a prominent cartilaginous rod within the forming jaw complex during the late embryonic stage of development.

In mammals, over many millions of years, the original jaw-opening joint became less prominent and was incorporated into the middle ear as the malleus (derivative of articular bone of the lower jaw) and the incus (derivative of the ancestral quadrate bone in the skull). The incus connects with the stapes (a derivative of the second pharyngeal arch). The tympanic ring , a neural crest–derived bone that surrounds and supports the tympanic membrane, is derived from the angular bone , one of the first-arch membrane bones that overlies the proximal part of Meckel’s cartilage.

Formation of the Palate

The early embryo possesses a common oronasal cavity. In humans the palate forms between 5 and 10 weeks to separate the oral from the nasal cavity. The palate is derived from three primordia: an unpaired median palatine process and a pair of lateral palatine processes ( Figures 14.8 and 14.9 ).

The median palatine process is an ingrowth from the newly merged nasomedial processes. As it grows, the median palatine process forms a triangular bony structure called the primary palate . In postnatal life, the skeletal component of the primary palate is referred to as the premaxillary segment of the maxilla . The four upper incisor teeth arise from this structure ( Figure 14.10 ).

Formation of the palate involves (1) growth of the palatal shelves, (2) their elevation, (3) their fusion, and (4) removal of the epithelial seam at the site of fusion. The lateral palatine processes, which are the precursors of the secondary palate , first appear as outgrowths of the maxillary processes during the sixth week. At first, they grow downward on either side of the tongue ( Figure 14.11 ). During week 7, the lateral palatine processes ( palatal shelves ) dramatically dislodge from their positions alongside the tongue and become oriented perpendicularly to the maxillary processes. The apices of these processes meet in the midline and begin to fuse.

Similar to other facial primordia, outgrowth of the palatal shelves involves ectodermal–mesodermal interactions and specific growth factors. Initially shh is active throughout the epithelium of the palatal shelves. It is critical for growth of the palatal shelves. While development progresses, FGF-7 in the neural crest mesenchyme on the future nasal side of the palatal shelves represses the activity of shh on that side ( Figure 14.12 ). On the oral side, FGF-10 signaling maintains the activity of shh. Both the transcription factor Pax-9 and shh stimulate the activity of Osr-2 (odd skipped related-2), which along with Msx-1 is an intrinsic regulator of mesenchymal proliferation within the palatal shelves. Generation of a sufficient mass of mesenchymal cells is critical for normal fusion of the palatal shelves.

Despite many decades of investigation, the mechanism underlying the elevation of the palatine shelves remains obscure. Swelling of the extracellular matrix of the palatal shelves imparts a resiliency that allows them to approximate one another shortly after they become dislodged from alongside the tongue. Research suggests that the rapid closure of the palatal shelves is accomplished by the flowing of the internal tissues, rather than by a reaction that approximates the closure of swinging doors.

Another structure involved in formation of the palate is the nasal septum (see Figures 14.8 and 14.11 ). This midline structure, which is a downgrowth from the frontonasal prominence, reaches the level of the palatal shelves when the palatal shelves fuse to form the definitive secondary palate. Rostrally, the nasal septum is continuous with the primary palate.

The oral cavity is lined with ectoderm, which during the period of palate formation is covered by a layer of peridermal cells (see p. 154). During much of early oral development, the anti-adhesive property of the periderm is important in preventing various components of the early oral cavity from adhering to one another. Such lack of adhesion is important in elevation of the palatal shelves. The rare mutations that interfere with periderm formation in the oral cavity result in the fusion of the palatal shelves with the tongue and/or the maxilla. Later in development, the presence of periderm would interfere with fusion of the palatal shelves, so the peridermal cells must be removed from the site of fusion (see later in the chapter).

At the gross level, the palatal shelves fuse in the midline, but rostrally they also join the primary palate. The midline point of the fusion of the primary palate with the two palatal shelves is marked by the incisive foramen (see Figure 14.10 ).

Because of its clinical importance, fusion of the palatal shelves has been investigated intensively. When the palatal shelves first make midline contact, each is covered throughout by a homogeneous epithelium. The first step in fusion is removal of the peridermal cells, from the sides of the site of fusion largely through apoptosis. The remaining epithelial cells then adhere to one another, forming a midline epithelial seam . During the process of fusion, however, the midline epithelial seam disappears. The epithelium on the nasal surface of the palate differentiates into a ciliated columnar type, whereas the epithelium takes on a stratified squamous form on the oral surface of the palate. Significant developmental questions include the following:

• 1.

  What causes the disappearance of the midline epithelial seam?

• 2.

  What signals result in the diverse pathways of differentiation of the epithelium on either side of the palate?

The disappearance of the midline epithelial seam after the approximation of the palatal shelves involves several fundamental developmental processes ( Figure 14.13 ). Most of the epithelial cells at the fusion seam undergo apoptosis and disappear. Others may migrate out from the plane of fusion and become inserted into the epithelial lining of the oral cavity. Still other epithelial cells undergo a morphological transformation into mesenchymal cells. Transforming growth factor-β3 (TGF-β3) is expressed by the ectodermal cells of the distal rim of the palatal shelves just before fusion and loses prominence shortly thereafter. It plays an important role in stimulating apoptosis of the epithelial cells at the fusion seam. In TGF-β3–mutant mice, the lateral palatal shelves approximate in the midline, but the epithelial seam fails to disappear, and the mice develop isolated cleft palate.

Experiments involving the in vitro culture of a single palatal shelf of several species have shown clearly that all aspects of epithelial differentiation (cell death in the midline and different pathways of differentiation on the oral and nasal surfaces) can occur in the absence of contact with the opposite palatal shelf. These different pathways of differentiation are not intrinsic to the regional epithelia, but they are mediated by the underlying neural crest–derived mesenchyme. The mechanism of this regional specification of the epithelium remains poorly understood. According to one model, the underlying mesenchyme produces growth factors that influence the production and regional distribution of extracellular matrix molecules (e.g., type IX collagen). The way these events are received and interpreted by the epithelial cells is unknown.

Formation of the Nose and Olfactory Apparatus

The human olfactory apparatus first becomes visible at the end of the first month as a pair of thickened ectodermal nasal placodes located on the frontal aspect of the head ( Figure 14.14A ). Similar to the formation of the lens placodes, the formation of the nasal placodes requires the expression of Pax-6 and the action of retinoids produced in the forebrain. In the absence of Pax-6 expression, neither the nasal placodes nor the lens placodes form. The nasal placodes originate from the anterolateral edge of the neural plate before its closure.

Soon after their formation, the nasal placodes form a surface depression (the nasal pits ) surrounded by horseshoe-shaped elevations of mesenchymal tissue with the open ends facing the future mouth (see Figure 14.5 ). The two limbs of the mesenchymal elevations are the nasomedial and nasolateral processes . Formation of the thickened nasal processes depends on the retinoid-stimulated production of FGF-8 , which stimulates proliferation of the mesenchymal cells within the nasal processes. The source of these retinoids is the epithelium of the nasal pit itself. Meanwhile, production of retinoids by the forebrain diminishes. As a consequence, the frontonasal prominence, which depends on forebrain retinoids for supporting proliferation of its mesenchymal cells, is reduced. While the nasal primordia merge toward the midline during weeks 6 and 7, the nasomedial processes form the tip and crest of the nose along with part of the nasal septum, and the nasolateral processes form the wings ( alae ) of the nose. The receding frontonasal process contributes to part of the bridge of the nose.

Meanwhile, regulated by Wnt ligands, the nasal pits continue to deepen toward the oral cavity and form substantial cavities themselves (see Figure 14.14 ). By 61⁄2 weeks, only a thin oronasal membrane , derived from the nasal fin, separates the oral cavity from the nasal cavity. The oronasal membrane soon breaks down, thereby making the nasal cavities continuous with the oral cavity through openings behind the primary palate called nasal choanae (see Figure 14.9 ). Shortly after breakdown of the oronasal membrane, however, the outer part of the nasal cavity becomes blocked with a plug of epithelial cells, which persists until the end of the fourth month. With the fusion of the lateral palatal shelves, the nasal cavity is considerably lengthened and ultimately communicates with the upper pharynx.

Similar to the other major sensory organs of the head, the epithelium of each nasal pit induces the surrounding neural crest mesenchyme to form a cartilaginous capsule around it. In a three-dimensionally complex manner, the medial parts of the nasal capsules combine with more centrally derived deep neural crest mesenchyme to form the midline nasal septum and ethmoid bones. The lateral region of the nasal capsule forms the nasal bones. During the third month, shelflike structures called nasal conchae form from the ethmoid bones on the lateral wall of the nasal cavity. These structures increase the surface area available for conditioning the air within the nasal cavity. Late in fetal life and for several years after birth, the paranasal sinuses form as outgrowths from the walls of the nasal cavities. A developing sinus is lined by respiratory mucosa and is able to grow out through the nasal capsule by means of local degeneration of the cartilage of the capsular wall. It then expands into the newly forming bone of the maxillary process. The size and shape of the paranasal sinuses have a significant effect on the form of the face during its postnatal growth period.

At 6 to 7 weeks, a pair of epithelial ingrowths can be seen in each side of the nasal septum near the palate. Developing as invaginations from the medial portion of the nasal placode, these diverticula, known as vomeronasal organs (see Figure 14.11B ), reach a maximum size of approximately 6 to 8 mm at around the sixth fetal month and then begin to regress, leaving small cystic structures. In most mammals and many other vertebrates, the vomeronasal organs, which are lined with a modified olfactory epithelium, remain prominent and are involved in the olfaction of food in the mouth or sexual olfactory stimuli (e.g., pheromones).

The dorsalmost epithelium of the nasal pits undergoes differentiation as a highly specialized olfactory epithelium (see Figure 14.14 ). Differentiation of the olfactory organ and the vomeronasal organ requires the action of FGF-8 , which is produced in a signaling zone that surrounds the nasal pit. Beginning in the embryonic period and continuing throughout life, the olfactory epithelium is able to form primitive sensory bipolar neurons, which send axonal projections toward the olfactory bulb of the brain. Preceding axonal ingrowth, some cells break free from this epithelium and migrate toward the brain. Some of these cells may synthesize a substrate for the ingrowth of the olfactory axons. Other cells migrating from the olfactory placode (specifically, the vomeronasal primordium) synthesize luteinizing hormone–releasing hormone and migrate to the hypothalamus, the site of synthesis and release of this hormone in adults. The embryonic origin of these cells in the olfactory placode helps explain the basis for Kallmann’s syndrome , which is characterized by anosmia and hypogonadotropic hypogonadism. Cells of the olfactory placode also form supporting (sustentacular) cells and glandular cells in the olfactory region of the nose. Physiological evidence shows that the olfactory epithelium is capable of some function in late fetal life, but full olfactory function is not attained until after birth.

Formation of the Salivary Glands

Starting in the sixth week, the salivary glands originate as solid, ridgelike thickenings of the oral epithelium ( Figure 14.15 ). Historically, the germ layer of origin of the primordia of the salivary glands has been difficult to determine, but recent research suggests that they are derived from ectoderm.

As with other glandular structures associated with the digestive tract, the development of salivary glands depends on a continuing series of epitheliomesenchymal interactions. The initial interaction begins with an FGF-10 signal from the epithelial primordium to the underlying neural crest mesenchyme. Then the mesenchyme begins to produce FGF-10, which induces proliferation of the salivary epithelial cells and is required for further development of the salivary gland. The outgrowing glandular epithelium contains a primary duct region capped by an epithelial endbud. Parasympathetic neurons, derived from bipotential Schwann cell precursors (see p. 236), respond to Wnts produced by cells of the primary duct by forming a parasympathetic ganglion around the primary duct. The parasympathetic nerves, in turn, produce vasoactive intestinal peptide , which promotes growth of the duct and formation of a lumen within it.

CLINICAL CORRELATION 14.1

MALFORMATIONS OF THE FACE AND ORAL REGION

Cleft Lip and Palate

Cleft lip and cleft palate are common malformations with an incidence of approximately 1 in 1000 births (cleft lip) and 1 in 2500 births (cleft palate). Numerous combinations and degrees of severity exist, ranging from a unilateral cleft lip to a bilateral cleft lip associated with a fully cleft palate. More than 500 syndromic forms of cleft lip/palate and over 250 genetic defects leading to these conditions have been described.

Structurally, cleft lip results from the lack of fusion of the maxillary and nasomedial processes. In the most complete form of the defect, the entire premaxillary segment is separated from both maxillae, with resulting bilateral clefts that run through the lip and the upper jaw between the lateral incisors and the canine teeth ( Figure 14.16 ). The point of convergence of the two clefts is the incisive foramen ( Figure 14.17B ). The premaxillary segment commonly protrudes past the normal facial contours when viewed from the side. The mechanism frequently underlying cleft lip is hypoplasia of the maxillary process. This prevents contact between the maxillary and nasomedial processes from being established.

Cleft palate results from incomplete or absent fusion of the palatal shelves (see Figures 14.16B and 14.17 ). The extent of palatal clefting ranges from involvement of the entire length of the palate to something as minor as a bifid uvula. As with cleft lip, the developmental basis for cleft palate is usually multifactorial. Some chromosomal syndromes (e.g., trisomy 13) are characterized by a high incidence of clefts. In other cases, cleft lip and cleft palate can be linked to the action of a chemical teratogen (e.g., anticonvulsant medications). Experiments on mice have shown that the incidence of cleft palate after exposure to a dose of cortisone is strongly related to the genetic background of the mouse. In humans, mutations of MSX1 are strongly associated with nonsyndromic cleft palate. The higher female incidence of cleft palate may be related to the occurrence of fusion of the palatal shelves in female embryos approximately 1 week later than in male embryos, thus prolonging the susceptible period.

The genetic, cellular, and molecular basis of normal palate closure is complex. Even closure of the anterior and posterior parts of the palate operates under different combinations of molecular interactions. This accounts for the high frequency of these conditions.

Oblique Facial Cleft

Oblique facial cleft is a rare defect that results when the nasolateral process fails to fuse with the maxillary process, usually resulting from hypoplasia of one of the tissue masses ( Figure 14.18A ). This cleft frequently manifests as an epithelially lined fissure running from the upper lip to the medial corner of the eye.

Macrostomia (Lateral Facial Cleft)

An even rarer condition called macrostomia ( Figure 14.18B ) results from hypoplasia or poor merging of the maxillary and mandibular processes. As the name implies, this condition manifests as a very large mouth on one or both sides. In severe cases, the cleft can reach almost to the ears.

Median Cleft Lip

Another rare anomaly, median cleft lip, results from incomplete merging of the two nasomedial processes ( Figure 14.18C ).

Cleft Mandible

Cleft mandible ( Figure 14.19 ), also very rare, results from the lack of merging of the distal ends of the mandibular arch. It has been reported only in patients with mutations of the EIF4A3 gene ( Richieri-Costa -Periera syndrome ).

Holoprosencephaly

Holoprosencephaly includes a broad spectrum of defects, all based on defective formation of the forebrain (prosencephalon) and structures whose normal formation depends on influences from the forebrain. This condition has been estimated to be present in as many as 1 in 250 of all embryos and 1 in 10,000 live births. The defect arises in early pregnancy when the forebrain is taking shape, and the brain defects usually involve archencephalic structures (e.g., the olfactory system). Because of the influence of the brain on surrounding structures, especially the cranial base, primary defects of the forebrain often manifest externally as facial malformations, typically a reduction in tissue of the frontonasal process.

In extreme cases, holoprosencephaly can take the form of cyclopia (see Figure 8.21 ), in which the near absence of upper facial and midfacial tissue results in a convergence and fusion of the optic primordia. Reduction defects of the nose can also be components of this condition. The nose can be either absent or represented by a tubular proboscis (or two such structures), sometimes even located above the eye (see Figure 8.21 ). Midline defects of the upper lip can also be attributed to holoprosencephaly (see Figure 14.18C).

The root cause of holoprosencephaly occurs very early in embryonic development, with disturbances in the ability of the prechordal plate and anterior endoderm to secrete sonic hedgehog (shh) and other factors required for induction and early development of the ventral forebrain. In their absence, the single optic field either does not split or splits incompletely, and ventral forebrain structures do not develop. This is also reflected in a reduced rostral neural crest, which provides the cellular basis for the formation of most midface structures. Even earlier in development, disturbances in bone morphogenetic protein (BMP) levels, often caused by imbalances in BMP inhibitors, can influence early formation of the forebrain and lead to holoprosencephaly.

Many cases of holoprosencephaly (e.g., Meckel’s syndrome , which includes midline cleft lip, olfactory bulb absence or hypoplasia, and nasal abnormalities) can be attributed to genetic causes. Meckel’s syndrome is an autosomal recessive condition. Several types of hereditary holoprosencephaly result from mutations of the shh gene, which normally induces the formation of several midline structures in the forebrain. Exposure to an excess of retinoic acid, which causes misregulation of genes in the shh pathway, also causes holoprosencephaly in laboratory animals and possibly in humans. Most cases of holoprosencephaly seem to be multifactorial, although maternal consumption of alcohol during the first month of pregnancy is suspected to be a leading cause of this condition. One percent to 2% of infants born to diabetic mothers may develop some degree of holoprosencephaly. Trisomies of chromosomes 13 and 18 are commonly associated with holoprosencephaly.

Frontonasal Dysplasia

Frontonasal dysplasia encompasses various nasal malformations that result from excessive tissue in the frontonasal process. The spectrum of anomalies usually includes a broad nasal bridge and hypertelorism (an excessive distance between the eyes). In very severe cases, the two external nares are separated, often by several centimeters, and a median cleft lip can occur ( Figure 14.20 ).

Overall, facial abnormalities cover a broad spectrum, ranging from profound tissue deficiencies (severe holoprosencephaly) to considerable excesses of tissue that lead to the partial duplication of the face ( Figure 14.21 ). They can be generally correlated with levels of shh activity, with low levels resulting in deficiencies and higher than normal levels associated with excess mid-facial tissue.

At the most severe end of the holoprosencephaly spectrum is cyclopia. A slightly less severe, but usually lethal form is manifest by two partly fused eyeballs in a single optic cavity and possibly a single nostril. Between these extreme conditions and normal are varieties of hypotelorism , characterized by reductions in nasal structures and a reduced space between the eyes. There is no sharp line between hypotelorism and what is socially perceived as normal. On the other side of the spectrum, at some point individuals are perceived to have more widely spaced eyes than normal. This condition is known as hypertelorism , and as the amount of midfacial tissue is increased, individuals can experience partially or completely separated external nares. As the tissue excess becomes more extreme, actual duplication of the nose or even the entire face ( diprosopus ) can occur.

Much of the development of an exocrine gland, such as the salivary glands, depends upon branching morphogenesis , which involves an interplay between the glandular epithelium and the extracellular matrix surrounding it. While the primary duct begins to mature, it forms a basement membrane around itself as a means of stabilizing the duct. Branching of an endbud involves a combination of proteolytic remodeling the basement membrane in the area where a cleft will form (see Figure 14.15 ) and cellular changes within the epithelium at the branch point. Overall expansion of an endbud occurs under the influence of FGF signaling coming from the surrounding mesenchyme. The response to the FGF signal is proliferation of the epithelial cells and migration of additional epithelial cells to the area of expansion. Expansion is facilitated through enzymatic digestion of the basement membrane, which becomes perforated and more distensible. The FGF signaling also prevents lumenization and duct formation in the endbud region. Cleft formation occurs as the result of apical constriction through local contractions of ordered microfilaments of the epithelial cells at the site of cleft formation, along with the deposition of fibronectin and collagen IV within the cleft itself (see Figure 14.15E ). As the endbud expands, the proximal region becomes more removed from the FGF influence, and small microlumens are formed within the epithelium. These will coalesce to form a continuous ductule, and the surrounding epithelial cells will constitute a branch of the overall duct system.

During organogenesis, the parasympathetic innervation, acting through acetylcholine secretion, maintains the population of epithelial progenitor cells. In its absence, the amount of budding of epithelial lobules is dramatically reduced. The structural and functional differentiation of the epithelium of the salivary gland continues throughout fetal life.

Clinical Correlation 14.1 presents malformations of the face and oral regions.